Automated vehicles are one of the most promising solutions for the greatest challenges of modern mobility: emission reduction, effective time management and comfort for modern citizens, efficient space management especially in urban areas and last but not least mobility and self-determination for user groups with limitations as well as ever increasing road safety.
To validate new functions of automated vehicles, very extensive and cost-intensive real world test drives are currently required. In light of the accelerated pace of innovation and the economic aspects of the automotive industry, the current procedure for validation is becoming increasingly inefficient and will foreseeably no longer be viable. The VVM project is committed to contribute to solving this problem.
The VVM project aims to develop test procedures and to provide frameworks and methods for the safety verification of automated vehicles. VVM is working on the use case of Urban Intersections and focuses on driving functions up to full automation of vehicles (SAE Level 4 and 5).
The VVM project builds on the results of the PEGASUS project, which addressed the identification and description of critical scenarios and their implementation in general test cases for highly automated vehicles (Level 3), using the example of the motorway pilot. VVM extends the PEGASUS method to Level 4 and Level 5 of automated driving in an inner city environment, using the example of an urban intersection.
The VVM project is also working on integrating the verification of safety as an integral part of the development process. Already during the development of components and subsystems, the testability, verification and validation are to be considered as design goals ("design for testability"). The resulting systems can then be tested hierarchically, which has the effect that when components are updated, not the entire system has to be tested again. This approach allows test requirements to fit a wide variety of combinations of components and to be implemented and combined reliably and efficiently.
The VVM project aims to establish a seamless chain of reasoning for the proof of safety and to demonstrate its feasibility. All possible tools from digital simulation to real world test drives are used as efficiently as possible. Efforts are minimised, cost and time efficiency is increased.
The VVM project investigates the complexity of real traffic situations for automated vehicles using the example of an urban intersection. Here, road users, in particular automated vehicles, are confronted with the greatest challenges involving crossing and turning traffic, traffic lights, a variety of road users such as pedestrians, cyclists, public transport with priority, etc.
On the basis of this use case, the causal relationships for urban driving will be determined and described, especially those that pose a traffic hazard. Why urban intersections? The hazards that occur at urban intersections are typical for urban traffic but also occur in other traffic situations, such as interurban driving. The identified causal relationships and methods based upon them, which will be developed in the project, therefore represent a much larger range of traffic situations. The methods for urban intersections are thus transferable to other traffic areas and use cases.
Dr. Mark Schiementz, BMW AG
(7/2019 – 6/2023)
Technische Universität Braunschweig
RWTH Aachen University
Continental Teves AG
Opel Automobile GmbH
AVL Deutschland GmbH
FZI Forschungszentrum Informatik
Robert Bosch GmbH
Concept to the Goal Structuring Notation (GSN) meta model
Qualitative system and test requirements defined
Refinement of system and test requirements
System and test requirements by probabilities
Methodology for verification and validation for the urban environment
Description of the safety proof to be performed